A Recyclable Thermoresponsive Catalyst for Highly Asymmetric Henry Reactions in Water

Abstract

The synthesis of enantiomerically pure chiral β-nitroalcohols is a crucial objective in asymmetric catalysis. In order to efficiently obtain such chiral products, we developed a series of thermoresponsive, oxazoline–copper catalysts (Cu^II-PN_xFe_yO_z) via sequential reversible addition–fragmentation chain transfer (RAFT) polymerization. These catalysts can self-assemble in water into single-chain nanoparticles (SCNPs) with biomimetic behavior, in which intramolecular hydrophobic and metal-coordination interactions generate a confined hydrophobic cavity. Comprehensive characterization by FT-IR, TEM, DLS, CD, CA, and ICP analysis confirmed the nanostructure and composition. When applied to the aqueous-phase asymmetric Henry reaction between nitromethane and 4-nitrobenzaldehyde, the optimal catalyst (2.0 mol%) achieved a quantitative yield (96%) with excellent enantioselectivity (up to 99%) within 12 h. Furthermore, the thermosensitive poly(N-isopropylacrylamide, NIPAAm) block enabled facile catalyst recovery through temperature-induced precipitation above its lower critical solution temperature (LCST). This work presents an efficient and recyclable biomimetic catalytic system, offering a novel strategy for designing sustainable chiral catalysts for green organic synthesis.

1. Introduction

Chiral β-nitroalcohols serve as pivotal building blocks in organic synthesis, particularly in the pharmaceutical and agrochemical industries, where their scaffolds are frequently found in bioactive molecules and natural products [1,2,3] (Figure 1). The nitro group in these compounds offers versatile chemical handles for further transformations into valuable chiral amines, amino acids, and other nitrogen-containing functionalities [4,5]. Chiral β-nitroalcohols can be reduced to amino alcohols, oxidized to nitro ketones, dehydrated to nitro alkenes, and undergo substitution reactions through various heteroatoms’ nucleophilic attacks. Consequently, the efficient and stereoselective synthesis of enantiopure β-nitroalcohols remains a significant pursuit. Traditional methodologies often rely on stoichiometric chiral auxiliaries or metal-catalyzed asymmetric reactions [6,7,8,9]. Among these, catalytic asymmetric conjugate addition represents one of the most atom-economical and efficient routes [10,11,12]. The catalytic methodologies for the asymmetric Henry reaction are predominantly categorized into three types: chiral metal complexes [13,14], organocatalysts [15,16], and biocatalysts [17,18,19,20]. Notably, chiral copper complexes, coordinated with diverse privileged ligands such as bisoxazolines [21,22], chiral diamines [23], amino alcohols [24], and Salen derivatives [25,26], represent the most extensively investigated and efficient catalytic system, owing to their high activity and excellent adjustability. Organocatalytic systems, including chiral (thio)urea and cinchona alkaloid derivatives, offer a complementary metal-free strategy [27]. Concurrently, enzyme-based catalysis is valued for its exceptional stereocontrol under mild reaction conditions. A survey of the literature confirms that copper-based catalysts have been the predominant choice in reported studies. A major challenge reported for copper-based catalysts lies in their difficult recovery and the requirement for an exogenous base under standard reaction conditions [28,29]. Thus, the development of a catalyst that can effectively be recovered, has a certain alkalinity, and can induce chirality is of crucial importance in addressing this issue. This approach not only effectively activates the substrate but also facilitates efficient catalyst recovery and reuse.

Despite their efficacy, homogeneous chiral copper catalysts often face challenges related to catalyst recovery, metal leaching, and poor stability, which hinder their practical and sustainable application. Immobilizing chiral metal complexes onto solid supports has been a widely adopted strategy to facilitate catalyst separation and reuse [30,31,32]. However, conventional heterogeneous catalysts frequently suffer from reduced activity and enantioselectivity due to limited accessibility of active sites and inefficient diffusion in non-aqueous media [33,34]. To address these limitations, there is a growing impetus to develop smart, responsive catalytic systems that operate efficiently in green solvents, particularly water [35,36]. The design of temperature-sensitive polymeric supports that can modulate their physicochemical properties (e.g., hydrophilicity/hydrophobicity) with temperature changes offers a promising avenue to enhance catalytic performance and substrate compatibility in aqueous environments.

Oxazoline, as a type of chiral ligand, can remarkably influence the performance of metal catalysts in asymmetric catalytic Henry addition reactions [37,38]. However, most studies involve the in situ complexation of oxazoline with copper acetate in organic solvents, which easily leads to environmental pollution and makes it difficult to effectively recover and reuse the catalyst [38]. If oxazoline can be polymerized under controlled conditions with the temperature-sensitive material NIPAAm (N-isopropylacrylamide) to prepare a single-chain polymer [39,40], and the metal copper can coordinate with the nitrogen atom in the oxazoline unit, it can cause the polymer to undergo single-chain folding to form a nanoreactor [41,42]. This method can prepare a catalyst that can activate nitromethane, and the catalyst structure is stable and easy to recover. In recent years, researchers have reported that ferrocene can activate aldehyde groups [43]. Based on this, combined with the temperature-sensitive oxazoline ligand, the preparation of a three-block multifunctional temperature-sensitive polymer is of great significance. The temperature-sensitive unit promotes the dissolution and recovery of the catalyst [44,45,46], the ferrocene unit activates the aldehyde groups in aldehydes [47,48], and the copper oxazoline activates the substrate nitromethane [49]. This type of catalyst can form a nanoreactor through self-folding of the monomeric chain in the asymmetric Henry addition reaction in the aqueous phase, concentrating the substrate in the nanoreactor cavity, promoting the efficient Henry reaction in the aqueous phase, exhibiting biomimetic catalytic characteristics, and enabling convenient recovery and repeated use of the catalyst.

Herein, by using the stepwise reversible addition–fragmentation chain transfer polymerization (RAFT) method, the thermosensitive unit NIPAAm was combined with the double-bond ferrocene unit for controlled polymerization to obtain a macromolecular chain initiator. Adding vinyl oxazoline to the macromolecular chain initiator for controlled polymerization resulted in the preparation of a series of three-block thermosensitive chiral catalysts. This catalyst can self-assemble into single-chain aggregated nanoreactors in water, providing a hydrophobic, confined reaction cavity for the asymmetric Henry reaction in the aqueous phase, demonstrating nanoreactor catalytic properties. The reaction only requires 2 mol% of the catalyst and can achieve a 4-nitrobenzaldehyde yield of 95% within 12 h, with an enantiomeric selectivity of up to 99%. For other electron-withdrawing group-substituted benzaldehyde substrates, certain catalysts also exhibit good catalytic effects. Moreover, due to the presence of the thermosensitive unit, after the reaction, the catalyst can be recovered and reused by heating, with a slight decrease in activity upon reuse. This work has, for the first time, constructed a water-phase asymmetric Henry addition reaction using the oxazoline system and achieved good catalytic effects.

2. Results and Discussion

2.1. Characterization of Materials Structure

The FT-IR spectrum of the conventional chiral oxazoline catalyst NC-Bn-Cu (Figure 2a) confirms its successful synthesis and metal coordination. Key features include the C=N stretch of the oxazoline unit at 1659 cm⁻¹, the oxazoline ring vibrations at 1059 and 982 cm⁻¹, the C-O-C stretches at 1267 and 1190 cm⁻¹, and a distinct Cu-N coordination band at 514 cm⁻¹. Figure 2b presents the spectrum of 4-vinylbenzyl ferrocene carboxylate, which shows the integrity of the ferrocene skeleton (peaks at 1124 and 1029 cm⁻¹) [50], a C=O stretch at 1713 cm⁻¹, and a C-H stretch at 3125 cm⁻¹, confirming the successful incorporation of the double-bond unit. The characteristic bands of the thermosensitive PNIPAAm block (Figure 2c) are identified as the -NH stretches of -CONH- at 3309 and 3072 cm⁻¹, -CH stretches at 2876 cm⁻¹ (for -CH(CH₃)₂), and 1029 cm⁻¹ (for -CH-CH-), alongside amide-related vibrations at 1653 and 1540 cm⁻¹ [51]. Significantly, these PNIPAAm features are retained in the final triblock polymer catalyst Cu^II-PN₅₀Fe₅O₅ (Figure 2d), verifying the presence of the thermosensitive unit. Furthermore, the catalyst spectrum exhibits a new band at 512 cm⁻¹, attributable to Cu-O coordination, indicating successful complexation of copper with the oxazoline nitrogen. Notably, the band at 1658 cm⁻¹ in the catalyst appears broadened, likely due to the overlap of the amide carbonyl stretch from PNIPAAm with the characteristic peak of the ferrocene unit.

The morphology of the catalyst in the aqueous solution was observed using transmission electron microscopy (TEM). As shown in Figure 3, the TEM image of the catalyst Cu^II-PN₅₀Fe₅O₅ in water is presented.

Owing to the hydrophilic nature of the thermosensitive PNIPAAm block, the catalyst Cu^II-PN_xFe_yO₅ exhibits complete water solubility at room temperature. In aqueous solution, it undergoes intramolecular self-assembly driven by synergistic hydrophobic and coordinative interactions, folding into well-defined single-chain nanoparticles (SCNPs). TEM analysis (Figure 3) confirmed the formation of uniform, monodisperse nanoparticles with an average size of approximately 2 nm. This structure constitutes a nanoreactor with a hydrophilic corona and a hydrophobic interior where the active catalytic sites are sequestered. In the asymmetric Henry reaction, hydrophobic aldehyde substrates and nitromethane were concentrated within these hydrophobic nanocavities. The confined space forces close and frequent collisions between the substrates and the embedded catalytic centers, significantly accelerating the reaction rate. This mode of action, involving substrate pre-concentration and enhanced local encounters, mimics the efficiency characteristic of enzymatic catalysis. The TEM particles of the three catalysts with different hydrophilic/hydrophobic ratios (Cu^II-PN₂₅Fe₅O₅, Cu^II-PN₅₀Fe₅O₅, and Cu^II-PN₁₀₀Fe₅O₅) showed little difference. This is because the TEM test was conducted on the dry sample of the material. After the moisture was dried, the PNIPAAm ends in the thermosensitive material would contract, resulting in a smaller particle size.

Dynamic light scattering (DLS) analysis at 25 °C (catalyst concentration: 0.5 mg mL⁻¹) revealed hydrodynamic diameters (D_h) below 10 nm for all catalysts (Figure 4), consistent with single-chain nanoparticle (SCNP) morphology. This self-assembly was driven by intramolecular hydrophobic and coordinative interactions: the hydrophilic PNIPAAm block ensures aqueous solubility, while the hydrophobic catalytic centers were sequestered within, forming a nanoreactor. A clear trend was observed in the SCNP sizes: Cu^II-PN₂₅Fe₅O₅ (4.8 nm) < Cu^II-PN₅₀Fe₅O₅ < Cu^II-PN₁₀₀Fe₅O₅ (8.6 nm). This progression indicated that a higher proportion of hydrophobic oxazoline units strengthens intramolecular folding, leading to a more compact conformation. All systems exhibited low polydispersity indices (PDI: Cu^II-PN₂₅Fe₅O₅ = 0.094, Cu^II-PN₅₀Fe₅O₅ = 0.083, and Cu^II-PN₁₀₀Fe₅O₅ = 0.121), confirming the formation of uniform and stable SCNPs. In contrast, the diblock analog Cu^II-PN₁₀₀O₅ formed larger nanoparticles (D_h = 13.7 nm) with a higher PDI. The comparatively larger and less uniform assembly of the diblock catalyst underscores the critical role of the terminal hydrophobic ferrocene unit in the triblock design. This unit enhances the overall hydrophobicity, which is the key driving force for forming compact, well-defined nanoreactors, thereby facilitating more efficient catalysis.

Based on TEM and DLS analyses confirming the formation of single-chain folded nanoparticles in water, the chirality of the catalyst Cu^II-PN₅₀Fe₅O₅ was further investigated by circular dichroism (CD) spectroscopy (Figure 5). Aqueous solutions (0.5 mg mL⁻¹) were examined across a temperature range from 288 K to 308 K (10 K intervals). The spectrum exhibited a negative Cotton effect centered at 259 nm. The intensity of this CD signal displayed a distinct, temperature-dependent trend. Initially, at 288 K, the Δε = −18.04 mdeg. The signal intensified progressively with increasing temperature, reaching a maximum value of Δε = −37.25 mdeg at 298 K (25 °C). This increase indicates an enhancement of the catalyst’s chiral environment. However, upon further heating past its lower critical solution temperature (LCST = 27.4 °C), the signal began to diminish. At 308 K (35 °C), where the polymer chain collapses due to hydrophobicity, the CD intensity weakened as catalyst solubility decreased. This profile demonstrates that the catalyst’s chirality is maximized just below its LCST [52]. Consequently, the asymmetric Henry reaction was conducted at 25 °C to leverage the optimal chiral induction provided by the catalyst.

We conducted a characterization of the LCST test for the catalyst (Figure 6). The results showed that when the solution temperature was below 27 °C, the solution exhibited a clear and transparent color, and the catalyst was hydrophilic. When the temperature was raised above 27.4 °C, the catalyst gradually became cloudy until it exhibited hydrophobic properties and precipitated out of the water. This is related to the properties of PNIPAAm. Below the LCST, Cu^II-PN₅₀Fe₅O₅ exhibited hydrophilicity. When above the LCST, the catalyst showed hydrophobicity. Therefore, the LCST of the Cu^II-PN₅₀Fe₅O₅ catalyst is 27.4 °C.

Meanwhile, the contact angle of the catalyst in water was 72° (Figure 7), indicating that it is hydrophilic at room temperature. This hydrophilic nanoreactor was also more conducive to the occurrence of catalytic reactions in the pure aqueous phase, thereby improving the catalytic reaction efficiency in the aqueous medium.

2.2. Catalytic Activity of Cu^II-PN₅₀Fe₅O₅ Catalyst

2.2.1. Optimization of Reaction Conditions

The catalytic performances of various copper catalysts were evaluated in the asymmetric Henry reaction of 4-nitrobenzaldehyde with nitromethane in water at 25 °C, with the key results summarized in

Table 1. Optimization of different catalysts in the Henry reaction.

Entry	Catalyst a	CH3NO2	Catalyst Amount (mol)	Time (h)	Yield b (%)	ee c (%)
1	Cu(CH3COO)2	3 mmol	2.0%	12	/	/
2	Cu(CH3COO)2	3 mmol	2.0%	48	4	/
3	NC-Bn-Cu	3 mmol	2.0%	12	20	35
4	NC-Bn-Cu	3 mmol	2.0%	48	31	35
5	CuII-PN50O5	3 mmol	2.0%	12	63	99
6	CuII-PN50O5	3 mmol	2.0%	18	92	99
7	CuII-PN25Fe5O5	3 mmol	2.0%	12	78	99
8	CuII-PN50Fe5O5	3 mmol	2.0%	12	96	99
9	CuII-PN100Fe5O5	3 mmol	2.0%	12	87	95
10	CuII-PN50Fe5O5	3 mmol	0.5%	12	14	99
11	CuII-PN50Fe5O5	3 mmol	1.0%	12	33	99
12	CuII-PN50Fe5O5	3 mmol	1.5%	12	58	99
13	CuII-PN50Fe5O5	3 mmol	2.5%	12	93	99
14	CuII-PN50Fe5O5	3 mmol	3.0%	12	93	99
15	CuII-PN50Fe5O5	1 mmol	2.0%	12	19	99
16	CuII-PN50Fe5O5	2 mmol	2.0%	12	65	99
17	CuII-PN50Fe5O5	4 mmol	2.0%	12	97	99

^a Catalyst (2.0 mol% of 4-nitrobenzaldehyde, based on copper content), 4-nitrobenzaldehyde (1.0 mmol), CH₃NO₂ (3.0 mmol), deionized water (1 mL), 25 °C. ^b Chemical yields were referred to isolated compounds and performed in triplicate. ^c Determined by HPLC (Daicel chiralpak AD column).

. The metal salt Cu(CH₃COO)₂ showed negligible activity (4% yield after 48 h). The conventional chiral oxazoline catalyst NC-Bn-Cu, hampered by poor water solubility leading to mass transfer limitations, afforded only 20% yield and a low 35% ee after 48 h (Entries 3, 4). In contrast, incorporating a thermosensitive PNIPAAm block dramatically improved performance. The diblock catalyst Cu^II-PN₅₀O₅ achieved 92% yield and 99% ee within 18 h, benefiting from its complete water solubility (Entries 5, 6). The triblock catalyst Cu^II-PN₅₀Fe₅O₅, which further self-assembles into single-chain nanoparticles (SCNPs) with a hydrophobic catalytic core, proved superior. It delivered a 78% quantitative yield and 98% ee in just 12 h (Entry 7). This enhancement was attributed to the hydrophobic ferrocene unit, which not only strengthens the nanoparticle’s compact structure but may also activate the aldehyde substrate. The hydrophilic/hydrophobic balance was found to be critical. Cu^II-PN₂₅Fe₅O₅, with a lower hydrophilic fraction, produced a lower yield (96% in 12 h, Entry 8), likely due to an overly constricted nanocavity that impedes the entry of substrate. Conversely, catalyst Cu^II-PN₁₀₀Fe₅O₅, with a higher hydrophilic fraction, formed a larger cavity that compromised stereocontrol, reducing yield to 87% and ee to 95% (Entry 9). Therefore, Cu^II-PN₅₀Fe₅O₅ represents the optimal design, delivering the highest activity and enantioselectivity.

The catalytic efficiency showed a clear dependence on catalyst loading (Table 1, Entries 10–14). At a low loading of 0.5 mol%, the yield after 12 h was merely 14%, indicating an insufficient number of active sites. Increasing the loading to 1.5 mol% raised the yield to 58%. A 96% yield was achieved only at 2.0 mol% loading, where the population of single-chain nanoparticle reactors is adequate for the substrate quantity. Further increases in catalyst loading did not enhance the yield, confirming 2.0 mol% as the optimal and economically viable point. Notably, the product yield and enantioselectivity remained consistently high (96%) across all loadings tested. The amount of nitromethane (CH₃NO₂) was also optimized (Entries 8, 15–17). With a stoichiometric amount (1 mmol, relative to 1 mmol aldehyde), the yield was poor (19%) after 12 h, likely due to its volatility over the prolonged reaction time. Using an excess (3 mmol) ensured sufficient CH₃NO₂ partitioned into the hydrophobic nanocavities, driving the reaction to near-quantitative yield (97%) with maintained high enantioselectivity (99% ee). A larger excess did not further improve the outcome. Therefore, 3 mmol of CH₃NO₂ was selected as the optimal condition, balancing complete conversion with atom economy. We innovatively designed a three-block polymer by incorporating ferrocene units to tune the composition and enable the self-assembly of well-defined nanocavities. DLS studies confirmed that upon copper coordination, the polymer folds into a nanoscale reactor with a confined cavity. The outermost temperature-sensitive block forms a hydrophilic shell, improving aqueous dispersion and reducing surface tension. The middle ferrocene segment enhances substrate activation and slightly boosts catalytic performance compared to the two-block analog. At the core, copper oxazoline serves as the active catalytic site, while the hydrophobic interior enriches hydrophobic substrates, promoting efficient contact with the catalyst. Overall, this design significantly accelerates the reaction and allows for convenient catalyst recovery and reuse via temperature-responsive precipitation.

2.2.2. Expansion of Reaction Substrates

The substrate scope of the asymmetric Henry reaction was evaluated using the triblock catalyst Cu^II-PN₅₀Fe₅O₅ under standardized conditions (25 °C, H₂O, 2.0 mol% catalyst, 3 mmol CH₃NO₂). The results (

Table 2. The asymmetric Henry reaction between nitromethane and aromatic aldehydes catalyzed by Cu^II-PN₅₀Fe₅O₅ in water.

Entry	Substrate	Product	Yield b (%)	ee c (%)
1			31	87
2			96	99
3			35	85
4			63	92
5			88	99
6			30	90
7			19	87

^a Catalyst (2.0 mol% of substrate, based on copper content), substrate (1.0 mmol), CH₃NO₂ (3 mmol), deionized water (1 mL), 25 °C, time = 12 h. ^b Chemical yields were referred to isolated compounds. ^c Determined by HPLC (Daicel chiralpak AD column).

) indicate that aromatic aldehydes bearing ortho-substituents generally performed better than benzaldehyde, with electron-deficient substrates exhibiting superior reactivity. Strong electron-withdrawing groups significantly enhanced the reaction. For instance, 4-nitrobenzaldehyde was consumed completely within 12 h, affording the product in >99% yield and 99% ee (Entry 2). This high efficiency is attributed to the increased electrophilicity of the carbonyl carbon. Similarly, 2-nitrobenzaldehyde achieved an 88% yield with 99% ee (Entry 5). 4-Cyanobenzaldehyde also showed high activity (63% yield, 92% ee, Entry 4). Substrates with halogen substituents, which are weakly electron-withdrawing, reacted more slowly. 4-Bromobenzaldehyde produced a 30% yield with 85% ee after 12 h (Entry 6). Notably, 2-chlorobenzaldehyde exhibited further diminished reactivity due to combined electronic and steric effects. Conversely, electron-donating groups markedly suppressed the reaction. 2-Methylbenzaldehyde, possessing an ortho-methyl group, delivered the poorest result with only a 19% yield (Entry 7), as the increased electron density on the aromatic ring reduces the carbonyl’s electrophilicity. In summary, the catalyst demonstrates high efficiency and enantioselectivity for electron-deficient aromatic aldehydes, with ortho-substituted variants showing particularly favorable outcomes in this aqueous nanoreactor system. Based on the experimental results, we believe that the possible catalytic mechanism involves the formation of a catalytically active dimeric complex. This catalytic particle with higher oxygen bridge basicity facilitates the reversible nitro–aldehyde–ketone reaction, thereby generating nitroalcohol [53,54].

2.3. Recycle and Reuse of Cu^II-PN₅₀Fe₅O₅

The catalyst Cu^II-PN₅₀Fe₅O₅ leverages the thermoresponsive property of its NIPAAm block for facile recovery. It is fully soluble in water at the reaction temperature (25 °C) but undergoes a sol-to-precipitate transition upon post-reaction heating, enabling its isolation by simple decantation. The recovered solid was washed, dried, and reused. However, a gradual decline in performance was observed over multiple cycles. In the aqueous asymmetric Henry reaction of 4-nitrobenzaldehyde, the second reuse led to diminished metrics (89% yield, 97% ee), which further decreased to 85%, 77%, and 75%, respectively, by the fourth cycle (Figure 8).

This deactivation was correlated with physical and spectroscopic changes. While the fresh catalyst formed a homogeneous dark blue solution, solutions of the recovered catalyst were lighter in color and semi-transparent, forming a slight precipitate upon standing. Comparative FT-IR analysis (Figure 2d vs. Figure 2d’) revealed a noticeable attenuation of the Cu-O coordination band in the reused catalyst. These observations collectively indicate a gradual leaching of copper ions during the recovery and reuse process, which is the primary cause of the observed activity loss. Nevertheless, this kind of catalyst exhibits superior efficiency and more convenient recovery than traditional catalysts in asymmetric Henry addition reactions.

3. Experiment

3.1. Materials

The following reagents were used as received: L-phenylalanine (AR, >99%), acryloyl chloride (AR, 97%), and p-toluenesulfonyl chloride (AR, >99%); N,N’-azobis(isobutyronitrile) (AIBN, AR, 98%), N-isopropylacrylamide (NIPAAm, AR, 99%), and sodium borohydride (AR, 99%). All other chemicals were laboratory-grade materials procured from local commercial sources. Solvents and reagents were purified following established standard methods where necessary.

3.2. Analytical Methods

Gel permeation chromatography (GPC) was performed on an Alltech system (Los Angeles, CA, USA) using THF as the eluent at a flow rate of 1 mL/min. Separations were carried out on a Jordi GPC 10,000 Å column (300 mm × 7.8 mm) equipped with an Alltech (Nicholasville, KY, USA) ELSD 800 detector, with the column and detector temperatures maintained at 30 °C and 40 °C, respectively. Polystyrene standards were used for calibration. Fourier transform infrared (FT-IR) spectroscopy was conducted on an AVATAR 370 Thermo Nicolet spectrophotometer (Shanghai, China). Samples were prepared as potassium bromide pellets, and spectra were recorded in the range of 400–4000 cm⁻¹ at a resolution of 4 cm⁻¹, averaging 32 scans. Dynamic light scattering (DLS) measurements were performed on a Malvern MS2000 Laser Particle Size Analyzer (Malvern, UK) to determine the hydrodynamic diameter. Transmission electron microscopy (TEM) images were acquired using an FEI Tecnai G20 microscope (FEI Tecnai G20, Hillsboro, OR, USA). For sample preparation, a droplet (~0.5 mg/mL) of the catalyst or polymer dispersion was placed on a carbon-coated copper grid and dried at room temperature for 72 h. The dried samples were then negatively stained with 1 wt% phosphotungstic acid and dried for an additional 72 h prior to imaging. UV-Vis spectroscopy was employed to determine the lower critical solution temperature (LCST) using an Agilent 8453 spectrophotometer (Shimazu UV-2700, Shanghai, China). Aqueous sample solutions (~1.0 mM) were monitored at 450 nm with a heating/cooling rate of 2 °C/min. Nuclear Magnetic Resonance (NMR) spectra (¹H and ¹³C) were recorded on a Bruker Avance-500 spectrometer(Rheinstetten, Germany) at 25 °C. Chemical shifts were referenced internally to tetramethylsilane (TMS; δ = 0.00 ppm for ¹H, δ = 77.00 ppm for ¹³C). The chiral properties of the catalyst were characterized by circular dichroism (CD) spectroscopy on a MOS-500 instrument (Grenoble, France). Spectra were recorded for an aqueous solution (0.5 mg mL⁻¹) at temperatures ranging from 278 to 308 K in 10 K increments.

3.3. Synthesis of Catalysts

3.3.1. Preparation of the Diblock Azolein Cu Catalyst Cu^II-PN₅₀O₅

Synthesis route of temperature-sensitive diblock azolein Cu catalyst (Cu^II-PN₅₀O₅): 2.5 mmol (0.3745 g) of the prepared 4-benzyl-2-allyl-4, 5-dihydrooxazoline monomer was weighed and dissolved in anhydrous methanol [55,56]. Then, 20 mmol of N-isopropylacrylamide NIPAAm (2.2632 g) was added into the above solution and transferred to a Schlenk tube. 1/6 mmol of thiophenyl acetic acid (a chain transfer agent, 0.0330 g) and 1/30 mmol of azobisisobutyronitrile (a chain initiator, 0.0052 g) were added to a Schlenk tube. After all the substances were completely dissolved, the solution was repeatedly evacuated and purged with N₂ until the entire solution was completely under a N₂ protection atmosphere. The reaction was carried out at 60 °C for 24 h. After the reaction, the reaction solution was vacuum concentrated, and the polymer was repeatedly precipitated with a multiple excess of ether to obtain a pale yellow product. It was then vacuum-dried at 30 °C to obtain a fluffy, light yellow triblock polymer product PN₅₀O₅ (Figure 9).

PN₅₀O₅: FT-IR (KBr): γ_max/cm⁻¹ 3424, 3267, 3062, 2969, 1651, 1416, 1360, 1275, 1231, 1211, 1128, 1034, 959, 915, 809, 683 cm⁻¹. The lower critical solution temperature (LCST) of the polymer PN₅₀O₅ was measured by UV spectroscopy and was found to be 31 °C. GPC (THF): M_n = 7159, M_w = 7341, PDI = 1.03. Then, 4 mmol of the prepared bimodal polymer PN₅₀O₅ was dissolved in 30 mL of anhydrous ethanol/ethyl acetate, and 2 mmol of copper acetate was added. The mixture was refluxed for 24 h. After the reaction, the solvent was evaporated, and a small amount of THF was added to dissolve it. Then, the solution was precipitated using multiple excess ether to obtain a blue solid precipitate. The precipitate was dried under vacuum at 30 °C to obtain the catalyst Cu^II-PN₅₀O₅. The catalyst was characterized by FT-IR (KBr): γ_max/cm⁻¹ 3423, 3062, 2965, 2930, 2875, 1728, 1614, 1556, 1454, 1395, 1336, 1272, 1174, 1135, 1105, 1044, 1016, 924, 836, 675, 617, 558, 511, 452 cm⁻¹. ${\mathsf{\alpha }}_D^{25}$ = −73.4 (C = 0.005 g mL⁻¹, CH₂Cl₂). The copper content was determined by titration: 0.304 mmol g⁻¹.

Synthesis of SCNPs type trimeric oxazoline Cu catalyst [57]: Take the 4-vinylbenzyl ferrocene carboxylate (2.5 mmol, 0.6924 g) prepared above, N-isopropylacrylamide NIPAAm (20 mmol, 2.2632 g) dissolved in anhydrous methanol, add it into the Schlenk tube, and add benzyl thioacetic acid (chain transfer agent, 1/6 mmol, 0.0330 g) and azobisisobutyronitrile AIBN (chain initiator, 1/30 mmol, 0.0052 g) (Figure 10). The solution was repeatedly evacuated and purged with nitrogen until the entire solution was completely under a nitrogen protection atmosphere. The reaction was carried out at 60 °C for 24 h to obtain the macromolecular chain transfer agent. At this time, the reaction was stopped, and 2.5 mmol (0.3745 g) of the third monomer (4-benzyl-2-viny-4, 5-dihydrooxazoline) was added, the air was vented, and nitrogen was purged again, and the reaction was continued at 60 °C for 24 h. After the reaction, the reaction solution was vacuum concentrated, and the product was precipitated using multiple excess ether, dried under vacuum at 30 °C, and obtained a fluffy, light yellow trimeric polymer product PN₅₀Fe₅O₅.

Maintaining the amounts of 4-vinylbenzyl ferrocene carboxylate and 4-benzyl-2-vinyl-4,5-dihydrooxazoline unchanged, varying the proportion of the temperature-sensitive unit NIPAAm monomer resulted in three different proportions of polymers PN₁₀₀Fe₅O₅, PN₅₀Fe₅O₅, and PN₂₅Fe₅O₅ (where x represents the polymerization degree of NIPAAm, y represents the polymerization degree of the ferrocene unit, and z represents the polymerization degree of the oxazoline unit). PN₁₀₀Fe₅O₅: FT-IR (KBr): γ_max/cm⁻¹ 3419, 3363, 3274, 3075, 2972, 2934, 2870, 1659, 1564, 1412, 1364, 1328, 1280, 1236, 1212, 1168, 1123, 1036, 952, 917, 805, 685, cm⁻¹. LCST = 29 °C. GPC (THF): M_n = 14307, M_w = 15175, PDI = 1.06. PN₅₀Fe₅O₅: FT-IR (KBr): γ_max/cm⁻¹ 3416, 3358, 3270, 3071, 2973, 2928, 2878, 1651, 1563, 1416, 1363, 1326, 1275, 1232, 1218, 1163, 1125, 1035, 946, 914, 807, 683 cm⁻¹. LCST = 28 °C. GPC (THF): Mn = 9025, Mw = 9164, PDI = 1.02. PN₂₅Fe₅O₅: FT-IR (KBr): γ_max/cm⁻¹ 3413, 3356, 3265, 3074, 2976, 2922, 2874, 1656, 1561, 1414, 1365, 1319, 1276, 1229, 1213, 1165, 1128, 1031, 945, 918, 803, 686 cm⁻¹. LCST = 27 °C. GPC (THF): Mn = 6164, Mw = 6269, PDI = 1.02.

3.3.2. Cu(II) Loading onto Supported Material Composite

Solutions of the three polymers were prepared in 25 mL of ethanol/ethyl acetate (1:1, v/v). Copper(II) acetate was added to each in a 2:1 molar excess with respect to the polymer’s binding sites, and the complexation was conducted under reflux for 24 h. Post-reaction, the solvent was evaporated off. The crude material was taken up in a minimal amount of THF and subjected to repeated precipitation from a large excess of diethyl ether. The blue solid complexes were isolated and then dried in vacuo at ambient temperature for 24 h, affording the desired catalysts: Cu^II-PN₁₀₀Fe₅O₅, Cu^II-PN₅₀Fe₅O₅, and Cu^II-PN₂₅Fe₅O₅, respectively. The characterization data of them are as follows: Cu^II-PN₁₀₀Fe₅O₅: FT-IR (KBr): γ_max/cm⁻¹ 3424, 3261, 3062, 2965, 2930, 2875, 1728, 1614, 1556, 1454, 1395, 1336, 1272, 1174, 1135, 1105, 1044, 1016, 924, 836, 675, 617, 558, 511, 452 cm⁻¹. ${\mathsf{\alpha }}_D^{25}$ = −69.1 (C = 0.005 g mL⁻¹, CH₂Cl₂). Cu content: 0.124 mmol g⁻¹ (see Supplementary Materials Determination of copper content). Cu^II-PN₅₀Fe₅O₅: FT-IR (KBr): γ_max/cm⁻¹ 3426, 3265, 3063, 2968, 2932, 2873, 1725, 1612, 1559, 1457, 1398, 1333, 1273, 1172, 1131, 1101, 1042, 1017, 922, 839, 673, 619, 559, 512, 453 cm⁻¹. ${\mathsf{\alpha }}_D^{25}$ = −68.4 (C = 0.005 g mL⁻¹, CH₂Cl₂). Cu content: 0.208 mmol g⁻¹. Cu^II-PN₂₅Fe₅O₅: FT-IR (KBr): γ_max/cm⁻¹ 3425, 3264, 3062, 2966, 2937, 2874, 1726, 1611, 1558, 1454, 1395, 1334, 1274, 1173, 1135, 1104, 1043, 1018, 925, 836, 674, 618, 553, 510, 464 cm⁻¹. ${\mathsf{\alpha }}_D^{25}$ = −67.7 (C = 0.005 g mL⁻¹, CH₂Cl₂). Cu content: 0.351 mmol g⁻¹.

3.4. Catalytic Activity Tests of Cu^II-PN_xFe_yO₅ Catalyst

3.4.1. General Procedure for Cu^II-PN_xFe_yO₅ Catalyzed Asymmetric Henry Reaction in Aqueous Phase

At 25 °C, the selected catalyst was completely dissolved in 1 mL of deionized water. To this solution, the aldehyde substrate (1 mmol) and nitromethane (CH₃NO₂, 3 mmol) were added sequentially under stirring. The reaction progress was monitored by thin-layer chromatography (TLC) using a petroleum ether/ethyl acetate mixture (4:1, v/v) as the eluent. Upon completion, the mixture was heated above the catalyst’s lower critical solution temperature (LCST), triggering its transition to a hydrophobic state and subsequent precipitation from the aqueous phase for easy recovery. The product-containing aqueous phase was then concentrated under reduced pressure and purified by column chromatography. The structures of the resulting chiral β-nitroalcohols were confirmed by ¹H NMR and ¹³C NMR nuclear magnetic resonance (NMR) spectroscopy, while their enantiomeric excess (ee) values were determined by high-performance liquid chromatography (HPLC).

3.4.2. Determination of Optimum Test Conditions

The optimization of reaction conditions for the Cu^II-PN_xFe_yO₅ catalyst involved screening key variables: copper loading levels; the identity and dosage of the chiral ligand; reaction time; and solvent effects. The efficiency of the asymmetric catalysis was subsequently quantified by measuring both the chemical yield and the enantioselectivity (ee value).

3.5. Recycling and Reusing of Cu^II-PN_xFe_yO₅

The Cu^II-PN_xFe_yO₅ catalyst demonstrated excellent recyclability over five cycles in the reaction with benzaldehyde. In a typical cycle, post-reaction workup involved ethyl acetate extraction and triple centrifugation at 6000 rpm to recover the solid catalyst. The recovered solid was washed with water and ethanol, then dried at 50 °C for 12 h before reuse. Inductively coupled plasma (ICP) analysis of the catalyst after the sixth cycle revealed no significant reduction in copper content, confirming its structural stability and correlating with the consistent catalytic performance observed throughout the recycling study.

4. Conclusions

(1)

A thermosensitive, single-chain triblock polymer (comprising NIPAAm as the thermosensitive unit, a ferrocene unit, and an oxazoline unit) was designed and synthesized via reversible addition–fragmentation chain transfer (RAFT) polymerization. This polymer was then coordinated with copper acetate to yield a class of biomimetic, single-chain aggregated chiral oxazoline–copper catalysts, denoted as Cu^II-PN_xFe_yO_z. For comparison, a control catalyst lacking the ferrocene unit, Cu^II-PN₅₀O₅, was also prepared to elucidate the role of ferrocene in the system. The structures of these catalysts and their self-assembled morphologies in water were confirmed through characterization by FT-IR, TEM, DLS, and CD.

(2)

This series of catalysts is fully water-soluble and undergoes intramolecular folding in aqueous solution to form single-chain nanoparticles, driven by hydrophobic interactions and metal coordination. As catalysis proceeds at 25 °C, the hydrophilic segments encapsulate the hydrophobic active sites, mimicking biological systems. The catalytic activity of the triblock catalysts was significantly higher than that of both a traditional chiral oxazoline–copper catalyst (NC-Bn-Cu) and the diblock control catalyst (Cu^II-PN₅₀O₅). Cu^II-PN₅₀Fe₅O₅ showed the highest activity, and the nanoreactor effect enhances its enantioselectivity.

(3)

While the catalyst (Cu^II-PN_xFe_yO₅) is soluble in water at 25 °C, it can be recovered post-reaction by leveraging its thermosensitive property. Raising the temperature of the reaction mixture triggers the precipitation of the hydrophobic catalyst as a solid, allowing for convenient separation. However, this recovery process was accompanied by the leaching of copper ions. Consequently, after four reuse cycles, both the yield and enantioselectivity showed a notable decrease. Thus, although the catalyst can be readily recovered, its reusability remains limited under the current post-treatment conditions.